Companies expect to bear more costs as a result of the autonomous systems they implemented badly.
Shapecharge | E+ | Gety pictures
Artificial intelligence capabilities develop rapidly, and companies worldwide are trying to keep pace with and implement artificial intelligence tools, but there are consequences for dirty implementation.
In fact, 79 % of companies expect the world to bear the “artificial intelligence debts” as a result of the independent tools that are implemented poorly, according to a new report submitted. Asana on the state of artificial intelligence at work Which surveyed more than 9,000 knowledge workers across the United States, the United Kingdom, Australia, Germany and Japan.
The report highlighted that companies are not prepared and lack the infrastructure and control required to enhance smooth cooperation between human employees and autonomy. Artificial intelligence agents. It differs from obstetric artificial intelligence, agents can act independently, and they can start procedures, and summon the previous work they have done. Some examples include Openai worker And human Clauds.
Artificial intelligence debts is the cost of not implementing the properly emerging independent systems.
“These costs can be money costs,” Hoffman said. “It may also be lost, which is related to money. This may also be a lot of things that you have to back away, which is costly from a financial point of view. It burns people to do so. It is all the costs associated with weak implementation,” Hoffman said.
The report explained that religion can be shown as security risks, poor data quality, low artificial intelligence agents that will waste time and resources for human employees, and the gap of management skills.
Hoffman said that this is not a comprehensive list and that “religion” could seem like a handful of code that was created by artificial intelligence that does not work properly or content created from artificial intelligence that no one uses.
A new research from Betterup Labs and Stanford Social Media Lab found that 40 % of the office workers have received “Worksloop” created by artificial intelligence, “ The researchers knew as content that looks good but lacks any substance.
Nearly two hours of additional work for the people they faced, an invisible tax worth $ 186 per month, and 9 million dollars have been increased productivity per year, according to the research.
“There is a major investment in this field at the present time, and in the end it is the issue of whether these investments will pay their fruits,” Hoffman said.
Henry Ajdir, founder of Consulting Consulting Consulting Consulting, Consulting Consulting Consulting, Consulting Consulting Consulting, Consulting Consulting Consulting, Consulting Consulting Consulting, Consulting Consulting Consulting, Consulting Consulting Consulting, Consulting Consulting Consulting, Consulting Consulting Consulting, Consulting Consulting Consulting, Founder of the UK Government, Video Starting, and AI, confirmed the need for implementation and studied structures.
“People who are from CTO or innovation officers, and the righteous people I have worked with, and who I think I have done the best position to succeed in it, they do not paint sugar that this will cost … as with any type of basic reformulation, you will face problems, you will get it up to the road.”
“It is not a magic silver bullet.”
The Asana report found that although artificial intelligence adopted to 70 % in 2025 of 52 % in 2024, workers also face higher levels of digital fatigue.
Digital exhaustion increased to 84 % in 2025 of 75 % in the previous year, while unmissible work burdens increased to 77 %, according to the report.
Moni Moreschid, CEO of Gill Gill, a US -based employment organization CNBC, told CNBC that although companies publish artificial intelligence tools and encourage their use, workers are still struggling.
“The primary reason is that they are struggling, and we know this from talking also to our graduates, is that the state of use of how and why this AI tool is supposed to use your work to flow is often missing.”
She added: “Without a clear understanding of what is the state of use that will make this task in particular better, faster and cheaper … This is what leads to fatigue, because you do not know the intended result.”
Mourshed pointed out that companies are investing in artificial intelligence in the hope that the work will be carried out overnight, faster and cheaper, but they do not provide training or instructions necessary to enable improvements.
“It is not a magic silver bullet, and suddenly do everything you want as soon as it is installed … it will be a much more painful journey to reach those benefits from the companies you thought about.”
Ajder intelligence expert said that the right strategy is the test of artificial intelligence and infrastructure building it instead of rushing to the race is not ready.
He said: “Do not start with inclusion only, you start with experimentation, and start through the range, through Sandboxing, by experimenting with these systems.”
This includes everything from the correct training of employees, to thinking about the type of artificial intelligence models that the work may need. It is difficult to respond to errors or breakdowns when there is no action in force.
“So I do not say that you cannot take care of carefully when it comes to the use of artificial intelligence, but it must be calculated and it must be determined,” said Ajar.
https://image.cnbcfm.com/api/v1/image/108204429-1758889385327-gettyimages-1427840896-pi-2631715.jpeg?v=1758901276&w=1920&h=1080
Source link